View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Research Papers in Economics

Crime and Town Centers: Are Downtowns More Dangerous Than Suburban Shopping Nodes?

Authors Richard Peiser and Jiaqi Xiong

Abstract The perception of high crime rates in downtowns has hindered the revitalization of downtown shopping districts and adjacent residential areas. This paper presents a better methodology for measuring crime in commercial shopping districts, replacing the conventional method of quoting crimes per 100,000 residences with a measure that more accurately reflects one’s chance for being a crime victim. This new measurement is used to address the question of whether the downtown shopping districts of Los Angeles and San Diego are as dangerous as two of their most competitive suburban shopping areas—Santa Monica and Fashion Valley. The findings indicate that actual crime rates in both downtown Los Angeles and downtown San Diego are in fact lower that those of their suburban counterparts.

Introduction

Shopping districts have historically anchored the core of downtown business districts. Since World War II, shopping in the United States has slowly but steadily migrated to suburban shopping malls. Flagship stores in downtown locations have gradually closed leaving in their wake downtown shopping districts struggling to attract shoppers back downtown from the suburbs. The long-term shift of retail activity from downtowns to suburban shopping malls has been addressed in numerous studies about retail investment and urban growth (e.g., for example, DiPasquale and Wheaton, 1996). These authors point to a number of factors that have contributed to downtown decline including traffic congestion, interstate highway development, government housing programs such as FHA, suburban housing, retail, office, and industrial development, and ongoing decentralization of jobs and housing. In addition, to these forces ‘pulling’ jobs and shopping out of downtown, they note ‘push factors’ including crime, poor public schools, and growing concentrations of minority and immigrant populations around downtowns. Shifts in shopping behavior from transit-oriented downtown department stores and main street shops to car-oriented regional shopping malls

JRER ͉ Vol. 25 ͉ No. 4 – 2003 578 ͉ Peiser and Xiong and big-box warehouse stores have also eroded the downtown shopping customer base over time. There are many signs that downtowns are coming back, but their role today is very different from what it was fifty years ago. Rather than being focal points for jobs and regional shopping, downtowns are coming back as cultural centers with significant numbers of residences and shopping that is often entertainment- oriented and must compete successfully with suburban counterparts. During the last decade, downtown shopping districts have been transformed in many cities to cater to new types of shoppers—tourists, immigrants and office workers. Many cities have built shiny new malls in downtowns trying to lure suburban shoppers back downtown. Great efforts have been made not only to reduce crime in downtown shopping districts but also to advertise urban regeneration efforts and bring people back from suburban shopping malls to shop downtown. As one of the important factors contributing to downtown decline and the persisting reluctance of shoppers to return downtown, this paper focuses on fear of crime. Previous studies have shown that perceptions of crime are slow to die and affect people’s shopping behavior and other location decisions long after the situation on the ground has changed. This study asks the simple question: Is shoppers’ perception of crime in downtown versus suburban shopping districts accurate? If not, how can downtown advocates overcome this misperception and help revitalize downtown retailing, assuming that they can overcome other problems retarding downtown recovery.1 The approach in this study is to measure the incidence of being the victim of a crime in two major metropolitan areas—Los Angeles and San Diego. The likelihood of victimization in the downtown shopping cores is compared with a major suburban competitor in each city. Shopping districts were deliberately chosen in high income areas for comparison—Santa Monica, a suburb of Los Angeles, and Fashion Valley, a suburb of San Diego. If downtown shopping centers are going to succeed, they must be competitive with the most attractive of their suburban counterparts. The analysis is complicated by the fact that victimization rates are normally measured by dividing the number of crimes by the number of residents in an area. This approach may work for comparing crime rates for residents across cities and suburbs, but it does not work for business districts that have low permanent residential populations but high daytime worker and nighttime visitor populations. The major problem of comparing crime rates, therefore, is how to measure the number of people subject to crime in a given geographic area. Surprisingly, the literature on crime rates in business districts is sparse, and other researchers’ solutions for dealing with this measurement problem are not readily available. This paper contributes to the literature in two ways. First, a more accurate methodology for measuring crime in commercial shopping districts is developed, which replaces the conventional method of quoting crimes per 100,000 residents. Crime and Town Centers ͉ 579

An attempt is made to measure the actual likelihood that an individual will be a crime victim based on the total number of people who frequent the shopping district—a number that includes shoppers, office workers, residents and tourists. Secondly, this new measurement is used to address the question of whether the downtowns of two major cities, Los Angeles and San Diego, are as dangerous as two of their most competitive suburban shopping areas. Reported crime rates influence people’s perceptions of crime. To preview the conclusions, by properly estimating one’s likelihood of being a crime victim in shopping centers and business districts, the findings demonstrate that downtown Los Angeles and San Diego have much lower crime rates than conventional measures would indicate. By comparing the rates with two of their wealthier suburbs, Santa Monica and Fashion Valley, the study shows that the central city downtowns are in fact safer than their suburban counterparts. The study builds on two principles established separately in the crime literature and real estate literature. In the crime literature, victimization rates should properly account for the number of potential victims, and in the real estate literature, fear of crime hurts real estate values and inhibits investment. By showing that perceptions of crime in downtown are no longer accurate, the paper provides evidence to mitigate fear of crime as a factor affecting downtown revitalization. The sample for this conclusion is small—two downtowns and two suburban centers. It would of course be desirable to compare many cities and many suburban shopping districts. The focus is on two metropolitan areas because proper analysis requires block-by-block tabulation of both crimes and people. Further study of other cities is needed to generalize the results. However, the results are of interest and lend support to the notion that downtowns are not as dangerous as many people believe. These findings, if borne out by other similar studies, are important for the long- term health of downtown areas and for bringing some of the most important customers—those who fled to the high income suburban shopping centers—back to downtown. The findings also offer hope for drawing more residents to live in downtown and surrounding neighborhoods, as well as encouraging investors to finance their new homes and shopping centers.

Background and Literature Review

This paper combines several different strands of literature that are important to understanding the relationship between crime, shopping behavior and downtown revitalization. The major strands include fear of crime versus actual crime, victimization, the relationship of crime to property values and to location decisions, shopping mall safety and crime rate measurement. While crime is but one of a number of factors affecting the recovery of downtown shopping, perceptions of crime linger long after an area has become safer. Perception of crime versus actual crime is addressed directly in a recent paper by

JRER ͉ Vol. 25 ͉ No. 4 – 2003 580 ͉ Peiser and Xiong

Myers and Chung (1998). They find that prior victimization explains some of the rather enormous perception of future victimization, but a sizable gap between perceived risk and actual risk remains. In communities where there is widespread knowledge of the victimizations or of the victims, perception of the risks of victimization may be inflated. Such factors as racial composition of the neighborhood and neighborhood instability often show larger independent impacts on fears or perceptions of crime than do actual victimization experiences. Another paper by Dominitz and Manski (1997) uses the Survey of Economic Expectations. One of their findings is that respondents substantially over-predict the risk of burglary. Similarly, a paper based on the 1993 Minnesota Crime Survey (Minnesota Planning, 1994) finds that the percentage of respondents who believe they are likely to be victims of violent crime in the next year is much higher than the percentage reporting they had been victims in 1992. The relative risk of becoming a crime victim (or the level of fear) depends on one’s age and gender, where one lives, with whom they associate and what they do to protect themselves. The results revealed substantial disparities in fear of crime versus actual victimization experiences. While 44.5% of respondents feared walking alone at night within a mile of their homes and 21% expected to be threatened or attacked in the coming year, only 4% had actually been attacked in 1992. The current paper extends the work of Dominitz and Manski (1997), Myers and Chung (1998) and the Minnesota Crime Survey (1994) by computing victimization rates using a new methodology. The current study investigates how actual crime risk, using new methodology, compares to perceived crime risk as measured by computations using the conventional methodology. To the extent that fear of crime is influenced by reported crime rates, these papers support the thesis that the likelihood of being a crime victim may be overstated in business areas with high daytime populations, which are not taken into account under conventional reporting methodology. To the extent that perceptions of high crime linger long after actual crime has been reduced, the effects of misleading reported crime rates may inhibit real estate investment and revitalization for longer periods than is necessary. Fear of crime is closely related to signs of community disorganization and decay (Meithe, 1995). For example, litter on streets, vandalism, run-down buildings, or few neighborhood communication networks increase individual’s perceived risks of criminal victimization and their subsequent fear of crime. Fear of crime also influences when people work and shop. Hamermesh (1999) looks at the time of crime. He explains why there is a lower propensity to work evenings in large U.S. metropolitan areas. He finds significant impacts of inter-area differences in homicide rates on work timing—higher homicide rates deter working in the evening and at night and shift it to the daytime. He relies on evidence presented by Skogan (1990) that expressed fear of crime is highly positively correlated cross- sectionally with actual crime rates. He argues that the magnitude of fear of crime depends in large part on how it is measured. Fear of crime is a multiplicative function of both perceived risk and the perceived seriousness of the offense. Crime and Town Centers ͉ 581

Skogan describes the relationship between crime-related experiences and fear of crime as being complex and inconsistent across studies. He finds that the empirical evidence on the impact of measures of neighborhood incivilities on individual’s fear of crime is inconclusive. The present paper builds on the work of these studies by examining the relationship between actual crime rates in areas of perceived higher risk—namely, downtowns—as compared to their suburban counterparts. While providing support for the disconnect between fear of crime and actual crime, the paper also focuses on how poor measurement of crime may feed continuing perceptions of high crime, especially in non-residential areas where traditional measures do not accurately portray one’s risk of being a victim of crime. While downtown business districts are anchored by office buildings, many suburban business districts are anchored by shopping malls. There is a growing literature on the subject of crime and shopping malls. Phillips and Cochrane (1988) dispel some popular myths that shopping centers have high levels of crime and violence. The introduction of accurate recording of incidents showed that incidents of nuisance, not crime, were the predominant problem and these incidents stemmed from a conflict of interest between shoppers, retailers and the young people who were using the centers as meeting places. In an article four years later, Nazel (1992:28) notes a growing concern among developers that malls are not as safe as they used to be. Developers of projects in larger markets are reporting much more serious incidents, including increases in car theft and vandalism, drug dealers organizing transactions, warring gangs strolling the malls and more violent attacks on customers. Nazel remarks, ‘‘In a sense, the mall is now a street, with a great deal of pedestrian traffic. It’s a natural target for pickpockets and professional thieves who live off heavy traffic.’’ Concerns about mall safety have led to sharply higher budgets for security. Front- page reports of shootings in a mall or gang activity can hurt sales there for years. Ironically, perceptions that suburban malls are not as safe as they used to be may help downtown shopping areas enjoy a renaissance. In another relevant article, DiLonardo (1997) looks at the financial implications of retail crime prevention. The present article adds to the literature on crime and shopping malls by carefully measuring the differences in crime rates between downtown malls and the prime alternative suburban shopping destinations. The issue of crime in shopping malls is part of a broader literature on crime and place, and crime prevention through environmental design. For a general background on this topic, see Jeffrey (1971), Newman (1972), Poyner and Webb (1991), Clarke (1992), Felson (1995, 1998), Rengert (1996), and Felson and Clarke (1997, 1998). Eck (1997) examines 99 interventions in 78 studies of crime having to do with crime at places. He provides an excellent overview of the types of crimes in different places including apartments and residences, retail stores, banks and money-handling places, public transportation, parking lots and garages, open public spaces and public coin machines.

JRER ͉ Vol. 25 ͉ No. 4 – 2003 582 ͉ Peiser and Xiong

Another relevant strand of literature focuses on the relationship between location decisions, crime and property values. Segelhorst and Brady (1984) analyze the effect of fear of crime on location decisions of urban-suburban residents. They focus on the existence of fear as an ‘undepletable detrimental externality’ that promotes the movement of people from central cities to suburbs. They find that this fear generates a suburban incentive to discriminate against minorities to prevent the spread of violent street crime to the suburbs. Thaler (1978) also discusses crime and property values. He develops a hedonic model with property values as the dependent variable to determine how much people are willing to pay to live in a neighborhood with a low crime rate. The results indicate that crime has a statistically significant negative impact on property values. These two papers provide evidence concerning how crime affects location decisions. They motivate the present paper by establishing a connection between fear of crime and people’s buying decisions, especially with respect to decision of shoppers to shop in downtown versus the suburbs. To the extent that crime reduces property values, it affects investor decisions on where to invest in new shopping centers. Lastly, the present paper addresses the issue of crime rates and how these rates are reported. Engstad (1975) compared the number of auto crimes and bar crimes (assault, disorderly conduct and violations of the liquor act) in small areas with hotels to the same crime counts in adjacent areas without hotels. He standardized the crime counts by the number of residents living in the areas and found an association between the presence of hotels and higher rates of crime per thousand people. Bueger, Cohn and Petrosino (1995: 249) discuss the nature of boundaries in dealing with whether to count police presence just outside the boundary of a public place. The surveyors applied the ‘‘If I were a mugger’’ rule: ‘‘If I were a mugger, would I be deterred from mugging someone here and now because of the presence of the police at the particular location?’’ If the answer were ‘‘yes,’’ the observer was to record that officers were present in the hot spot even if they were physically outside the boundaries. These papers help provide an understanding of recording crime incidents—the numerator—for purposes of computing victimization rates. Articles that address the choice of denominators for standardizing crime counts are sparse. One exception is Eck and Weisburd’s article (1995:10) in which they discuss Engstad’s (1975) use of different denominators for auto and bar crimes. In his examination of crime near hotels, Engstad standardized crime counts in each hotel area by the number of parking places and by the number of bar seats in each area. He found that one particular hotel area had higher auto and bar crime rates than the other hotel areas. Another paper by Balkin and McDonald (1981) addresses the time that people are exposed to criminals. Their model distinguishes the ‘‘real’’ crime rate (the probability of being a victim per unit of exposure time) from the ‘‘nominal’’ crime rate (the number of crimes per capita). They note that the nominal crime rate can be inversely related to the real crime rate. More accurate reporting and accounting of crime is important if fear of crime is to be Crime and Town Centers ͉ 583

reduced in areas such as business districts where conventional reporting may overstate actual victimization rates. Taken together, the literature has a number of strands that are important to the present paper. Most of these articles are written from the perspective of criminology or sociology. This paper adds to this literature from a real estate perspective. By comparing downtown shopping center crime to that in suburban centers, this paper takes advantage of the several strands of literature noted above to evaluate the actual crime rates for competing shopping destinations.

͉ Methodology To address the question of whether downtown shopping centers have higher crime rates than suburban shopping centers, the likelihood of becoming a victim in the downtown shopping districts was compared with that in the suburban shopping district. Two metropolitan areas, Los Angeles and San Diego were selected for this research. For each city, a major high-end suburban shopping center was also selected—Santa Monica’s Third Street Promenade near the ocean, twelve miles west of downtown Los Angeles, and the Fashion Valley area, approximately eight miles northwest of downtown San Diego (see Exhibits 1 and 2). A high-end shopping area in each metropolitan area was selected for comparison purposes because well-to-do shoppers are the ‘bread and butter’ of shopping centers. It is their preference for suburban shopping malls that has motivated most major department stores to leave downtown areas. Successful suburban shopping centers in Santa Monica and Fashion Valley represent major competitive alternatives to downtown shopping areas in downtown Los Angeles and San Diego. While shoppers are drawn to suburban shopping malls for many reasons besides fear of crime downtown, by comparing downtown crime rates with those in the competitive suburban shopping nodes, insight can be gained into how far downtowns have to go with respect to crime to draw back the most favored group of shoppers. Presumably, if crime rates are such that shoppers who frequent the most upscale shopping centers can be drawn back downtown, then less well-to- do shoppers in other parts of town will be even more likely to do so.2 The actual crime occurrence can be measured by dividing the total annual Part I reported crimes to the total exposed population. Part I crimes, shown in the Appendix, include major crimes such as murder, rape, robbery, arson and burglary. The result gives the victimization rate3—the likelihood that one will become a crime target. The lower the victimization rate, the safer the place. The numerator in the function is the annual aggregated Part I reported crimes for the target area.4 The denominator is the total population exposed to crime in the target area. This includes anyone who would have some likelihood of being present in the area, including residents, employees, shoppers and visitors. Each group is aggregated on an annual basis.

JRER ͉ Vol. 25 ͉ No. 4 – 2003 584 ͉ Peiser and Xiong

Exhibit 1 ͉ Downtown Los Angeles & Santa Monica Third Street Promenade

Downtown Los Angeles

Santa Monica Third Street Crime and Town Centers ͉ 585

Exhibit 2 ͉ San Diego Downtown and Fashion Valley

Fashion Valley

Downtown San Diego

JRER ͉ Vol. 25 ͉ No. 4 – 2003 586 ͉ Peiser and Xiong

Victimization Rate ϭ Total Annual Crime Incidents/Total Exposed Population. (1)

Where:

Total Exposed Population ϭ (Residents) ϩ (Employees) ϩ (Shoppers) ϩ (Visitors). (2)

Time of the day in crime data is critical since the crime incidents that should be included in the numerator are the ones committed during hours that the stores are open, such as from 8:00AM to 8:00PM. Conceptually, crimes committed after closing should not be part of the calculation since they do not affect shoppers and other daytime workers and visitors. However, if an area has a high crime rate after stores close, it might still create a negative safety image to shoppers even though the crime rate during store hours is low. A shopper may be less likely to visit a shopping center where several midnight murders are reported if he or she has other alternative destinations. The denominator—the total crime exposed population—may be somewhat over- estimated due to the overlap among groups, illustrated in Exhibit 3. Residents may

Exhibit 3 ͉ The Total Crime Exposed Population

Denominator Residents Visitors

Overlaps

Shoppers

Employees Crime and Town Centers ͉ 587 work in the same area. Residents may also represent part of the shoppers, as may employees. Among the four groups, ‘‘visitors’’ is the only category that is, to a large extent, independent of the other three groups. Taking into account the overlaps between residents, employees and shoppers may increase the pool of the risk-exposed population. Identifying the extent of the overlaps is not practical due to the lack of information. However, since denominators in both downtown and suburbs are subject to similar overlaps, this study assumes that the potential errors in the Total Exposed Population are on the same order of magnitude and therefore do not alter the results.5 The Victimization rates are computed for two geographic areas in each city and suburb. The larger area—called the Study Area—encompasses the entire downtown core or suburban business district. The smaller area—called the Target Area—encompasses the immediate shopping district where major shopping destinations in the downtown or suburban town center are located. The Study Areas and Target Areas in Los Angeles and Santa Monica are shown in Exhibits 4 and 5. The Study and Target Areas San Diego and Fashion Valley are shown in Exhibit 6.6 Four comparisons were conducted based on the matrix shown in Exhibit 7 in order to gain insight into victimization rates in the four locations. For each comparison, the area that has a lower victimization rate is safer. Comparison 1 contrasts the victimization rates between the Study Areas and Target Areas for downtown Los Angeles and San Diego. Comparison 2 compares the Study Areas and Target Areas for each of the two suburbs, Santa Monica and Fashion Valley. These two comparisons provide insight into whether the shopping district Target Areas have higher crime rates than the entire downtown Study Areas in which they are situated. Comparison 3 contrasts the victimization rates for the Target Areas between the downtown and the suburb—Los Angeles Downtown versus Santa Monica, and San Diego Downtown versus Fashion Valley. Comparison 4 contrasts the Study Areas between the downtown and suburb for each metropolitan area. These two comparisons go to the heart of the study, providing the answer for the question: Are the downtown shopping areas and surrounding study areas more dangerous than their suburban counterparts? While the comparative crime rates in the shopping districts themselves are the primary focus for bringing people back downtown, crime rates of the downtown as a whole versus the suburb are also important. A shopping destination may be considered safer than its suburban counterpart if it has a lower victimization rate. However, if its surrounding area is not, the shopping destination may still suffer from the negative image of its broader neighborhood context.

JRER ͉ Vol. 25 ͉ No. 4 – 2003 588 ͉ Peiser and Xiong

Exhibit 4 ͉ Downtown Los Angeles Target Area vs. Study Area

Arco Plaza

7th Mkt Place

Macy’s Plaza

Study Area

Target Area

͉ Data

Area Selection

Initially, six metropolitan areas around the country were examined. Los Angeles and San Diego were chosen because data was more readily available and their proximity and our familiarity with them made it easier to define the boundary areas. The boundary areas were based on input from the police and planning Crime and Town Centers ͉ 589

Exhibit 5 ͉ Santa Monica Target Area and Study Area In BEAT Map Format Overlaid with Census Tracts

departments in the four areas, and by the natural boundaries of the areas. The western boundary of the Study Area in downtown Los Angeles coincides with the boundary for the Target Area, for example, because the shopping district is on the west side of downtown and extends to the Harbor Freeway. This freeway forms a natural boundary both for the downtown core (the Study Area) and for the Target Area (see Exhibit 4). Similarly, the southern and western boundaries in downtown San Diego are co- terminus for both the Study Area and the Target Area because the shopping district is on the ocean (see Exhibit 6). The Study Area for the Los Angeles downtown is the core central business district encircled by freeways. This area, which is approximately 1.5 miles by 2 miles includes not only the downtown office, retail, government, and performing arts cores, but also the old garment district and skid row. The Target Area is the newer downtown shopping district. It contains the indoor and outdoor shopping malls of Macy’s Plaza, 7th Street Market place and Arco Plaza. This area is the major

JRER ͉ Vol. 25 ͉ No. 4 – 2003 590 ͉ Peiser and Xiong

Exhibit 6 ͉ Downtown San Diego and Fashion Valley Area Target Area and Study Area

Fashion Valley Mall

Pacific HWY

Downtown SD Study Area Downtown SD Target Area Fashion Valley Study Area Fashion Valley Target Area

Horton Plaza

shopping district catering to tourists, office workers and suburban shoppers (see Exhibit 4)7 Santa Monica is an independent municipality contiguous to Los Angeles approximately 10 miles west of downtown on the Pacific Ocean. While it has many wealthy residents and is adjacent to some of the most expensive parts of the city, it also has a sizeable renter, elderly and immigrant population. Third Street Promenade in Santa Monica—a pedestrian street in the heart of the shopping district—has become one of the most popular shopping destinations for tourists as well as residents throughout west L.A.8 San Diego has seen one of the most successful revivals of downtown shopping in the U.S. through the redevelopment of Horton Plaza and the Gas Lamp District Crime and Town Centers ͉ 591

Exhibit 7 ͉ Victimization Rate Comparisons

(A) (B) Downtown Suburb

Comparison 1

LADTA SMDTA (1) Shopping Destinations SDDTA SDFVTA 3 Target Area 2 Comparison Comparison LADSA SMDSA (2) CBD/Town Centers SDDSA SDFVSA Study Area

Comparison 4

Abbreviations: Target Areas: LADTA—Los Angeles Downtown Target Area

SMDTA—Santa Monica Downtown Target Area SDDTA—San Diego Downtown Target Area SDFVTA—San Diego Fashion Valley Target Area Study Areas: LADSA—Los Angeles Downtown Study Area SMDSA—Santa Monica Downtown Study Area SDDSA—San Diego Downtown Study Area SDFVSA—San Diego Fashion Valley Study Area

over the last twenty years. While it is a much more popular tourist destination than downtown Los Angeles—in part because it is located next to the ocean—the downtown shopping district must compete with numerous suburban shopping centers. The Study Area is defined by freeways and natural boundaries; the Target Area, by police BEAT areas, which contain the shopping district around Horton Plaza (Exhibit 6).9 San Diego’s Fashion Valley area contains the largest concentration of major suburban shopping centers within the City of San Diego, approximately 8 miles northwest of downtown at the conjunction of two major freeways. It is located in an area that is traditionally the wealthier part of town. The police department’s Western Division includes a very large area extending to the ocean. While the

JRER ͉ Vol. 25 ͉ No. 4 – 2003 592 ͉ Peiser and Xiong

Target Area for Fashion Valley was clearly defined by the immediate shopping district, the entire area of the Western Division was initially included in the analysis. Subsequently, a smaller area was defined, which excluded the ‘Peninsula area’ west of Pacific Highway. While the Peninsula area is part of the Western Division, it is cut off from the rest of the study area by Pacific Highway.10 All four areas include major indoor shopping malls as well as street-oriented retail shops. San Diego’s Horton Plaza and Santa Monica’s Third Street Promenade provide open-air shopping experiences that are particularly attractive to tourists and young people. While both downtown San Diego and Los Angeles have seen extensive redevelopment over the last twenty years, San Diego has been more successful at creating a downtown shopping environment that is likely to draw residents from other parts of town.

Crime Data

The data for calculating victimization rates consists of two primary data sets: (1) Part I crime data as the numerator; and (2) the number of residents, employees, shoppers and visitors in the specified Study and Target Areas as the denominator. For the numerator, Part I crime data collected from local police departments in Los Angeles, Santa Monica and San Diego was used. The Part I crime data was collected from local police departments by BEAT as of year-end 1997. Part I crime data cover the entire spectrum of violent crimes that create the most significant safety concerns in an area (see the Appendix). For example, crime data for both the Study and Target Areas in downtown Los Angeles were obtained from the 1997 Los Angeles Police Department Selected Crimes and Attempts Report, BEAT of Central Division. The Study Area covers 44 BEAT reporting units. The Target Area where major shopping malls are located contains reporting units 151, 152, 161 and 162. The locations of the reporting districts for these areas are shown in Exhibit 4. A similar procedure was used for the other three geographic areas. These maps and detailed crime data tallies are available from the authors.

Residents, Employees, Shoppers and Visitors

For the denominator, resident statistics were obtained from the U.S. Census Bureau. In each Study Area and Target Area, the total number of residents was aggregated by census tracts that corresponded with the crime reporting districts. Since census tract boundaries do not always conform to the crime data BEAT boundaries, adjustments needed to be made for accuracy. The methodology used in making adjustments assumes that population distributes evenly in each census tract. As such, if a census tract is larger than the boundaries defined by the crime BEAT, the total number of residents is reduced by the percentage of area outside the boundary. For example, the number of residents in 1990 for census tract 2100 Crime and Town Centers ͉ 593 in downtown Los Angeles is 5,552. However, since approximately 50% of the census tract 2100 is located outside of the Target Area boundary, 50% of the residents were included in the aggregation process. In addition to the above adjustment, 1990 census data was converted to 1997 figures. In downtown Los Angeles and Santa Monica, a countywide average growth rate of 5% over seven years was applied, using information from the California Department of Finance, Demographic Research Unit. A countywide growth rate of 11.9% was applied to downtown San Diego and the Fashion Valley Area based on information from the San Diego Association of Governments. The number of employees and shoppers for all four areas was taken from the 1997 Daily Trip Data provided by the Southern California Association of Governments and the San Diego Association of Governments. The Trip Data identifies employees and shoppers whose travel destinations are inside the Target and Study Areas. As such, employees are those traveling from various places to ‘‘work’’ while shoppers are estimated from the number of trips to ‘‘shop’’ in the Target and Study Areas. The number of overnight visitors in downtown Los Angeles and Santa Monica was approximated by multiplying the total hotel room inventory by the 1997 annual hotel occupancy rates provided by the local Convention and Visitor Bureau for each city. The study assumed 1.57 average occupancy per occupied room.11 Also, the number of overnight visitors was assumed to be the same for both the Target Area and the Study Area on the grounds that hotels in the Study Area generate the visitors who shop in the Target Area. Most downtown hotels are located near the Target Area but not inside it. It seems reasonable to assume that hotel guests in the downtown area are likely to shop in the main downtown shopping area. Therefore, Study Area hotel guests were included in the computation of the Total Exposed Population (Equation 2). The Total Exposed Population thus includes everyone who lives, works, shops, or visits the Study and Target Areas during the daytime. A breakdown of crimes by time of day was not available for all the areas studied, so crime data includes crimes committed during nighttime.12 Exhibit 7 shows the breakdown for residents, workers, shoppers and visitors, the sum of which gives Total Daytime Population for downtown Los Angeles and Santa Monica. Exhibit 8 shows the resident population and Total Daytime Population for downtown San Diego and Fashion Valley. Exhibits 9Ð11 summarize the crime incidents and victimization rates for the four areas studied, broken down by type of crime. These tables provide results for Comparisons 1 and 2, which contrast the Target Areas with the Study Areas for each of the four geographic areas. Exhibit 10 shows the crime counts and victimization rates for downtown Los Angeles. The victimization rates in the Target Area are well below those of the

JRER ͉ Vol. 25 ͉ No. 4 – 2003 594 ͉ Peiser and Xiong

Exhibit 8 ͉ Statistics of Daytime Population Downtown Los Angeles and Santa Monica

Total Daytime Residents Workers Shoppers Visitors Population

LADTA 655 62,055 977 4,258 67,946

LADSA 24,672 217,070 8,693 4,258 254,693

SMDTA 2,457 36,878 1,660 3,725 44,720

SMDSA 20,806 55,618 4,840 3,725 84,988

Notes: Worker and shopper data are extracted from SCAG trip data. Resident data is obtained from 1990 U.S. Census and California Department of Finance, Demographic Research Unit. The 1990 resident data is adjusted by a 5% growth rate from 1990 to 1997. (LA County population growth rate is close to 5% from 1995 to 97: In 1997, the estimated population is 9,456,000; in 1990, it is 8,902,000).

Exhibit 9 ͉ Statistics of Daytime Population San Diego Downtown and Fashion Valley

Total Daytime Residents Population

SDDTA 16,021 72,322

SDDSA 86,238 159,178

SDFVTA 15,852 21,998

SDFVSA 102,163 113,396

Notes: Daytime population and resident data are provided by SANDAG. SANDAG daytime population is calculated based on trip data. However, the data does not contain a breakdown of residents, workers, shoppers and visitors. Crime and Town Centers ͉ 595 Property Crime Violent Crime Target Area vs. Study Area, by Type of Crime Downtown Los Angeles 1997 Daytime Part I Crime Victimization ͉ Agg. Assault Burglary B/T Auto Pers Theft Other Theft Auto Theft Exhibit 10 Murder Rape Robbery Daytime population is used to calculate the victimization rate. Panel A: Crime Incidents LADTALADSA 13Panel B: Victimization LADTA 27 2LADSA 0.00% 0.00%Note: 918 0.00% 82 0.01% 0.12% 641 0.36% 41 0.06% 0.25% 715 0.20% 137 0.28% 1,525 0.28% 189 0.60% 0.04% 273 0.11% 29 0.63% 1,764 0.69% 0.05% 425 385 0.15% 0.18% 37 1,600 0.63% 1.20% 4,662 1.83% 125 817

JRER ͉ Vol. 25 ͉ No. 4 – 2003 596 ͉ Peiser and Xiong

Study Area, both for violent crime and for property crime (Comparison 1). This confirms what one would expect since the Study Area includes Skid Row east of Main Street but the Target Area does not.13 Theft and burglary dominate the crime categories. Exhibit 11 shows the breakdown of crime for Santa Monica. A comparison of victimization rates (Comparison 2) shows that the Target Area around Third Street Promenade is somewhat more dangerous than the Study Area (the rest of Santa Monica’s downtown). This is the case for all types of crime except residential burglary and vehicle theft. Since the rest of downtown Santa Monica is a mixture of low-to-mid-rise office buildings, hospital buildings, minor retail, and moderate- density residential buildings, and there is no area comparable to Skid Row in downtown Los Angeles, it is not surprising that the Target Area has slightly higher crime than the rest of downtown. Exhibit 12 shows the results for downtown San Diego. The downtown Target Area has about half the crimes of the Study Area. It has relatively higher theft, commercial burglary, and robbery, and relatively lower crimes in the other categories for murder, rape, aggravated assault and residential burglary. Like downtown Los Angeles, downtown San Diego has a Skid Row area, although it is smaller than Los Angeles’. Also, unlike Los Angeles, there are more high- income residents who live downtown. Victimization rates for the downtown San Diego Target Area around Horton Plaza (Comparison 1) are similar to those of the Study Area. Rates are slightly lower for violent crime in the Target Area, and slightly higher for property crime than in the Study Area. The breakdown of crime for San Diego’s Fashion Valley is shown in Exhibit 13. For 1997, 131 violent crimes and 1,482 property crimes occurred in the Target Area. The great majority of crimes are theft and vehicle theft. As the population figures in Exhibit 9 show, both the Fashion Valley Target Area and Study Area are predominantly residential in character. Because the Study Area is mostly residential, it includes many more residential burglaries. Also, robbery and aggravated assault are surprisingly high in the Study Area. Exhibit 13 shows the comparative victimization rates (Comparison 2) for San Diego’s suburban Fashion Valley Study and Target areas. Victimization rates for violent crime are lower in the Target Area than the Study Area (0.59% vs. 1.05%) but property crime victimization rates are higher in the Target Area (6.74% vs. 5.64%). The crime rate is surprisingly high in the Fashion Valley area—double that of downtown.

Comparison of Downtown vs. Suburban Shopping Districts

Exhibits 14 and 15 answer the question: Are downtowns more dangerous than their suburban counterparts? These exhibits show the results for Comparisons 3 Crime and Town Centers ͉ 597 Property Crime Violent Crime Vehicle Theft Arson Commercial Burglary Theft Residential Burglary Target Area vs. Study Area, by Type of Crime Santa Monica Downtown 1997 Daytime Part I Crime Victimization ͉ Agg. Assault Exhibit 11 Murder Rape Robbery Daytime population is used to calculate the victimization rate. Panel A: Crime Incidents SMDTASMDSAPanel B: Victimization SMDTA 7SMDSA 0.00% 14 0.00%Note: 0.02% 86 126 0.02% 0.19% 0.15% 122 60 0.13% 0.14% 0.01% 68 5 0.08% 0.21% 0.17% 144 92 2.95% 2.18% 1,853 0.21% 1,319 0.26% 0.01% 224 0.00% 0.34% 94 0.31% 4 3.38% 3 2.70% 262 153 2,293 1,513

JRER ͉ Vol. 25 ͉ No. 4 – 2003 598 ͉ Peiser and Xiong Property Crime Violent Crime Vehicle Theft Commercial Burglary Theft Residential Burglary Target Area vs. Study Area, by Type of Crime San Diego Downtown 1997 Daytime Part I Crime Victimization ͉ Agg. Assault Exhibit 12 Murder Rape Robbery Daytime population is used to calculate the victimization rate. Panel A: Crime Incidents SDDTASDDSA 2 7Panel B: Victimization SDDTA 18SDDSA 55 0.00% 0.00%Note: 174 0.02% 358 0.03% 0.24% 243 0.22% 730 0.34% 0.46% 56 304 0.08% 0.19% 171 0.24% 341 0.21% 1,565 2.16% 2,692 1.69% 0.41% 295 0.48% 763 0.60% 436 0.72% 1,150 2.89% 2.58% 2,087 4,100 Crime and Town Centers ͉ 599 Property Crime Violent Crime Vehicle Theft Commercial Burglary Theft Residential Burglary Target Area vs. Study Area, by Type of Crime Agg. Assault San Diego Fashion Valley Area 1997 Daytime Part I Crime Victimization ͉ Exhibit 13 Murder Rape Robbery Daytime population is used to calculate the victimization rate. Panel A: Crime Incidents SDFVTASDFVSA 1 8Panel B: Victimization SDFVTA 4 36SDFVSA 0.00% 0.01%Note: 0.02% 45 368 0.03% 0.20% 0.32% 784 81 0.37% 0.69% 709 0.16% 36 0.63% 0.35% 353 0.31% 77 4.28% 3,984 3.51% 942 1.95% 1,351 1.19% 428 0.59% 1,196 1.05% 6.74% 131 6,397 5.64% 1,482

JRER ͉ Vol. 25 ͉ No. 4 – 2003 600 ͉ Peiser and Xiong

Exhibit 14 ͉ 1997 Victimization Comparison Between Downtown LA and Santa Monica

Downtown LA Santa Monica

Approach LADTA (%) LADSA (%) SMDTA (%) SMDSA (%)

Proposed 1.39 2.46 3.73 3.01 Conventional 143.77 25.38 67.81 12.28

and 4. These exhibits also show the conventional calculation of victimization rates based solely on residential population. Exhibit 14 compares the victimization rates for downtown Los Angeles and Santa Monica. The conventional computation indicates that downtown Los Angeles is much more dangerous than suburban Santa Monica—143.77% versus 67.81% in the Target Areas, and 25.38% versus 12.28% in the Study Areas. The proposed computation, however, shows just the reverse—that downtown Los Angeles is safer than Santa Monica. Both the downtown Los Angeles Target and Study Areas have lower victimization rates than Santa Monica—1.39% versus 3.73% for the Target Areas (Comparison 3); 2.46% versus 3.01% for the Study Areas (Comparison 4). The proposed victimization rates offer a much better estimate of one’s likelihood of being a crime victim during the daytime than the conventional computation. While the daytime victimization rate for Santa Monica is more than double that of downtown Los Angeles in the Target Area, the conventional rate does provide an approximation of the nighttime victimization rate. Indeed, that rate is much higher in downtown Los Angeles than Santa Monica, both in the Target and Study Areas. The conventional victimization rate of 144% suggests that each resident will be victimized, on average, 1.5 times per year—a frightening possibility. The rate is so much higher in the Target Area because there are so few residents there.

Exhibit 15 ͉ 1997 Victimization Comparison Between Downtown San Diego and Fashion Valley

Downtown SD Fashion Valley SD

Approach SDDTA (%) SDDSA SDFVTA (%) SDFVSA (%) (%)

Proposed 3.49 3.30 7.33 6.70 Conventional 15.75 6.09 10.17 7.43 Crime and Town Centers ͉ 601

No doubt, these high rates are likely to impact consumers’ perceptions of the general safety in downtown. Nevertheless, the results indicate that such figures do not properly reflect the victimization risk to shoppers during daytime. It should be noted that as shopping centers cater more successfully to nighttime customers through restaurants and cinemas, shopping center owners have a stake in more careful accounting about the time and location of crime. Exhibit 15 shows that the difference between the proposed computation and the conventional computation for San Diego is even more dramatic. The conventional rate suggests that the downtown San Diego Target Area is more dangerous than Fashion Valley—15.75% versus 10.17%. In fact, the proposed computation indicates that the victimization rate in the downtown Target Area is less than half the rate in the suburban Fashion Valley Target Area—3.49% versus 7.33%. The difference, of course, is due to the fact that the conventional rate is based on a very small nighttime population whereas the proposed victimization rate is based on the far larger daytime population. Conventional rates suggest that the downtown Study Area is much safer than the downtown Target Area (6.09% versus 15.75%) because the Study Area boundaries include several residential neighborhoods around downtown. The proposed computation shows that the downtown Study and Target Areas have similar victimization rates (3.30% versus 3.49%). For the Study Areas, conventional rates show that downtown San Diego is somewhat safer than Fashion Valley—6.09% versus 7.43%. The proposed computation for the Study Areas, however, shows that victimization rate for downtown compared to Fashion Valley is significantly lower than the conventional rate suggests—3.30% versus 6.70%. A comparison of Los Angeles with San Diego reveals that crime rates in San Diego are higher than Los Angeles both in downtown and in the selected suburb. The crime rate for Fashion Valley was in fact so high that the analyses were questioned and rerun. One possible explanation for the surprisingly high victimization rate in Fashion Valley is that it is close to areas that are crowded and have significant poverty and deteriorated housing. While this is also true of Santa Monica, the size and proximity of such neighborhoods to Fashion Valley may have greater impact on the comparisons. Fashion Valley was selected in the first place based on recommendations by the San Diego Police Department and real estate developers familiar with the San Diego area who indicated that Fashion Valley was the dominant suburban shopping district. Nevertheless, Fashion Valley clearly is not located in as exclusive an area as that surrounding Santa Monica.14

͉ Conclusion Downtown shopping districts have been in decline since the 1950s when suburban shopping centers first started attracting shoppers away from them. Eventually, the department stores where they shopped moved away as well. Since then, downtown

JRER ͉ Vol. 25 ͉ No. 4 – 2003 602 ͉ Peiser and Xiong

landscapes have been littered by countless attempts—mostly unsuccessful—to bring shoppers back downtown. While many factors have contributed to the declining role of downtown shopping districts, this paper has focused on crime, and whether or not reported crime rates properly reflect one’s likelihood of being a crime victim. One of the principal impediments to the revival of downtown shopping districts is the perception that the areas have high crime rates and therefore are more dangerous than their suburban counterparts. This study has examined the crime data for two cities, Los Angeles and San Diego, in an attempt to see if the perceptions of high crime rates downtown are still justified. Changing the perceptions about crime are important for changing consumer attitudes about the attractiveness of alternative shopping destinations, and therefore to tenants and investors. Crime data is typically reported in terms of crimes per 100,000 residents. Such data is misleading for commercial shopping cores since the residential populations are small. This is especially true of downtown shopping cores that have even fewer residents than their suburban counterparts. Victimization rates are intended to report the likelihood that one will be a victim of a crime in a given area. The error embedded in computing victimization rates based on residents was corrected in a business district by measuring the daytime populations, which include not only residents but also workers, shoppers and out-of-town visitors. Conceptually, one’s likelihood of being a crime victim depends on the number of people in a given location at a given time. While the data is messy, the resulting numbers give a better measure of a person’s being a crime victim in the subject areas. The results for both downtown Los Angeles and downtown Santa Monica indicate that daytime victimization rates are higher in the suburban shopping district than in downtown. The same results hold for downtown San Diego versus suburban Fashion Valley. In every case, the crime rates are lower downtown than in the suburban comparison area, both for the Target Areas and the Study Areas. Careful tabulation of crime and population for Los Angeles and San Diego support the notion that at least these two downtowns are in fact safer than people think. Similar research conducted in other cities may add support to these results. In any event, to the extent that fear of crime has deterred investment in downtown areas, changes would appear to be occurring that make downtown areas safer and therefore more attractive for investment. ͉ Appendix ͉ Part I Crimes Murder Rape Robbery Aggravated Assault Crime and Town Centers ͉ 603

͉ Appendix (continued) ͉ Part I Crimes Burglary Residential Commercial Other Theft Vehicle Theft Arson Other

Data Source: Los Angeles Police Department, Santa Monica Police Department and San Diego Police Department.

͉ Endnotes 1 One of the reviewers provided this succinct description for the purpose of the paper. 2 This discussion presumes that crime rates in the wealthiest parts of town are as low or lower than other areas that are not as well-off. While criminals certainly are drawn to wealthier parts of town, residents of those areas typically have the political power to obtain the best police protection. They also can afford private security services as evidenced by the large number of homes in Santa Monica with yard signs indicating that the homes are guarded by private security patrols. 3 The victimization rate may have different definitions in other literatures. 4 The definition of Part I crime may vary between police departments. See the Data section. 5 Since residents and employees are population groups that present themselves in an area on a daily basis while shoppers and visitors present themselves at much lower frequencies, the victimization rate calculated through Equation (1) may underestimate the actual risk exposure of these two groups. However, since this issue is consistent between downtown and suburbs, both rates are assumed to be scaled down by the same factor, therefore, this issue does not challenge the accuracy of the comparison. 6 More detailed maps for downtown San Diego and Fashion Valley are available from the authors. 7 Another shopping district is located on the eastern side of downtown along Broadway. It includes the old movie theaters, mostly defunct, and is now a popular Latino shopping destination. 8 One of the reviewers questioned the choice of Santa Monica because it had its own pre- existing downtown before the mall was built. While the selection of a ‘‘greenfields’’ mall in Los Angeles for comparison may have led to different results, the Santa Monica mall is one of the most popular shopping destinations and clearly represents a prime shopping alternative to downtown. (Also see Endnote 13.) 9 BEAT is not an abbreviation but rather police lingo for a police patrol area or district.

JRER ͉ Vol. 25 ͉ No. 4 – 2003 604 ͉ Peiser and Xiong

10 The authors were least familiar with the suburban shopping area of San Diego and relied primarily on advice from the San Diego Police Department for determining the boundaries of the Fashion Valley Target and Study Areas. 11 The fast assistance of Doug Geoga and Penny Pritzker is gratefully acknowledged. 12 The inclusion of nighttime crime is likely to make victimization rates in the Study Areas somewhat higher relative to the Target Areas than they would otherwise be since the Study Areas include some blocks with concentrations of poverty and homelessness that the Target Areas do not. Nighttime crimes in these blocks are especially high. 13 Both the downtown Los Angeles and San Diego Study Areas include Skid Row areas with larger concentrations of homeless and single room occupancy hotels. In both cities, these areas are several blocks away from the Target Area, but their proximity influences perceptions of higher crime in the Target Area. 14 Santa Monica has lower income neighborhoods close to the Study Area in Venice and along Lincoln Avenue. One of the reviewers expressed concern about the selection of Fashion Valley since it was not in an independent suburb like Santa Monica. In terms of distance from the downtown shopping core and location in the wealthier part of town, it is comparable to Santa Monica. Fashion Valley was selected through an unbiased process. We concur with the reviewer that selection of another suburban shopping center such as one in La Jolla, might have presented a less dramatic case in favor of downtown. ͉ References Balkin, S. and J. F. McDonald, The Market for Street Crime: An Economic Analysis of Victim-Offender Interaction, Journal of Urban Economics, 1981, 10, 390Ð405. Bueger, M. E., E. G. Cohn and A. J. Petrosino, Defining the ‘‘Hot Spots of Crime’’: Operationalizing Theoretical Concepts for Field Research, In J. E. Eck and D. Weisburd (Eds.), Crime and Place, Crime Prevention Studies, Vol. 4, Criminal Justice Press, Monsey, N.Y. and The Police Executive Research Forum, Washington, D.C, 1995, 237Ð58. Clarke, R. V., Situational Crime Prevention: Successful Case Studies, New York, NY: Harrow and Heston, 1992. DiLonardo, R., Financial Analysis of Retail Crime Prevention, In M. Felson and R. Clarke, (Eds.), Business and Crime Prevention, 1997, 249Ð62. DiPasquale, D. and W. Wheaton, Urban Economics and Real Estate Markets, Englewood Cliffs, NJ: Simon and Schuster, Prentice Hall, 1996. Dominitz J. and C. Manski, Perception of Economic Insecurity, Public Opinion Quarterly, 1997, 61, 261Ð87. Eck, J., Preventing Crime at Places, In S. Lawrence, D. Gottfredson, D. Mackenzie, J. Eck, P. Reuter and S. Bushway (Eds.), Preventing Crime: What Works, What Doesn’t, What’s Promising, A Report to the United States Congress, Washington, DC: National Institute of Justice: 1997. Eck, J. and D. Weisburd, Crime Places in Crime Theory, In J. E. Eck and D. Weisburd, Crime and Place, Crime Prevention Studies, Vol. 4, Criminal Justice Press, Monsey, NY and The Police Executive Research Forum, Washington, DC, 1995. Engstad, P. A., Environmental Opportunities and the Ecology of Crime, In R. A. Silverman and J. J. Teevan (Eds.), Crime in Canadian Society, Toronto, Canada: Butterworths, 1975. Crime and Town Centers ͉ 605

Felson, M., How Buildings Can Protect Themselves Against Crime, Lusk Review for Real Estate Development and Urban Transformation, 1995, 1:1, 1Ð7. ——., Crime and Everyday Life, Second edition, Thousand Oaks, CA: Pine Forge Press, 1998. Felson, M. and R. V. Clarke, (Eds.), Business and Crime Prevention, Monsey, NY: Criminal Justice Press, 1997. ——, (Eds.), Crime Prevention in Rental Apartments, Security Journal, Vol. 11, 1998. Hamermesh, D. S., Crime and the Timing of Work, Journal of Urban Economics, 1999, 45, 311Ð30. Jeffrey, C. R., Crime Prevention Through Environmental Design, Beverly Hills: Sage, 1971. Meithe, T. D., Fear and Withdrawal from Urban Life, The Annals of the American Academy of Political and Social Science, 1995, 539, 14Ð27. Minnesota Planning, Troubling Perceptions: 1993 Minnesota Crime Survey, Minnesota Criminal Justice Statistical Analysis Center, St. Paul, Minnesota, 1994, 12Ð5. Myers, S. L. Jr., and C. Chung, Criminal Perceptions and Violent Criminal Victimization, Contemporary Economic Policy, 1998, 16:3, 321Ð33. Nazel, D., Crime in the Malls: A New and Growing Concern, Chain Store Age Executive, 1992, 68, 1992, 27Ð9. Newman, O., Defensible Space: Crime Prevention Through Urban Design, New York, NY: Macmillan, 1972. Phillips, S. and R. Cochrane, Crime and Nuisance in the Shopping Centre: A Case Study in Crime Prevention, Crime Prevention Unit: Paper 16, London: Home Office, 1988. Poyner, B. and B. Webb, Crime Free Housing, Oxford: Butterworth, 1991. Rengert, G., The Geography of Illegal Drugs, Boulder, CO: Westview, 1996. Segelhorst, E. and M. Brady, A Theoretical Analysis of the Effect of Fear on the Location Decisions of Urban-Suburban Residents, Journal of Urban Economics, 1984, 15, 157Ð71. Skogan, W. G., Disorder and Decline, New York, NY: The Free Press, 1990. Thaler R., A Note on the Value of Crime Control: Evidence from the Property Market, Journal of Urban Economics, 1978, 5, 137Ð45.

The authors are grateful to two anonymous reviewers for very detailed and helpful comments, and to Marcus Felson for introducing the authors to new areas of the crime literature on crime and place.

Richard Peiser, Harvard University, Cambridge, MA 02138 or rpeiser@gsd. harvard.edu. Jiaqi Xiong, Glenborough Realty Trust, San Mateo, CA 94043 or Jachi.Xiong@ glenborough.com.

JRER ͉ Vol. 25 ͉ No. 4 – 2003